Person A receives an allegedly unsolicited email from you.
Person A complains to Spam Cop.
Spamcop tells you "someone complained. we're getting you shut down."
You say "What was his email address? I have the logs of the date, time, and IP address from which he asked to receive this newsletter. If through some never-before-seen miracle he got on my mailing list without signing up for it, I'll be more than happy to take him off."
Spamcop says, "Sorry, that would make it too easy to prove that it's this idiot's fault for asking to receive special offers."
You say "fine, I'll just remove him from the list. What's his email address?"
spamcop says, "sorry, we'd much rather stop all bulk emailing than actually solve individual's problems. FOAD."
You missed the very important part about the complaint being sent from an alias to the email address of the complainer. SpamCop does not mediate in your hypothetical transaction. The spammer and/or the ISP both have the opportunity to reply to this user. If the user does not give a valid response, in this case including the original address, the ISP and spammer have no reason to change a thing. SpamCop can't automagically shut people down, and ISPs generally only consider taking action after getting large numbers of SpamCop and other abuse reports.
If you run a newsletter, some flakey users quickly forget they signed up for it, start complaining via SpamCop, and then never reply to your response to their contant spam reports, you can contact SpamCop and request to refuse reports with the email address removed or refuse SpamCop reports alltogether. SpamCop will even block users who abusivly submit reports of questionable quality.
It's true that local delivery is the MDA's job, but a powerful MDA is built in to Postfix to take care of this via maildir, mbox, pipe or LMTP (for handing off to Cyrus, for example). Sendmail, however, requires an external MDA and seperate configuration to do what can be configured in a base install of Postfix. That's what's great about Postfix - it isn't just an MTA, but a mail system. Best of all, its features are well integrated but seperate processes to make them highly efficient.
I agree. I'm guessing that at least some of these complaints were emails forwarded to uce@ftc.gov, which receives millions of emails. The FTC has to sift through these, figure out which ones are worth going after, then figuring out which ones are from the same spammer, and finally which emails are usefull in a court case.
Re:this SMTP server vs Qmail and Sendmail
on
Postfix 2.1 Released
·
· Score: 3, Informative
It is also important to note that Postfix provides Maildir support for local delivery. This means you can have nested folders (containing both messages and more folders) on your IMAP server, where as with Sendmail's mbox format you can only have folders containing messages, and those folders are actually just long text files. Qmail provides the maildir format natively, but Postfix makes it free.
this is, as far as i recall, the first time that apple has offred a video card option in its laptops.
It is not an option. If you click on one of the links so generiously provided in the post, such as the PowerBooks link, you can see that the option is either get a 15" or 17" PowerBook with an ATI Mobility Radeon 9700 or a 12" PowerBook with an NVIDIA GeForce FX Go5200.
Re:What about CPU/internet speed?
on
Paid To Spam
·
· Score: 1
Surely someone with a fast CPU on a fast connection is worth more per "CPU hour" than someone with a 486 on dialup?
The program probably is limited by IO, so the process sleeps once the outbound mail queue fills up. Translation: if you can only send 100 mails per hour on a Pentium 200 because of your bandwith limitation, the process will use just as much CPU as if the Pentium 200 were sending 100 emails in a few seconds over a 1.5Mbps link. However, you're right that an hour of CPU time on a 486 is worth much less than an hour of CPU time on an Athlon FX-53.
I don't know if anyone's tried scaling it. I imagine that for densely populated places that demand very large amounts of drinkable water on a daily basis, it would require impossible amounts of evaporation surface.
According to this page, the low end of the scale is about 11kWhr per 1,000 gallons (3,785L) for reverse osmosis. The Tampa plant produces up to 111,000,000 gallons per day. So that comes to 1,221,000 kWhr per day. We can skip the electrical conversion and use solar heat to evaporate and distill the water, but that requires much more power than reverse osmosis. I'm going to have to say this means it won't scale well, except perhaps in desert locations where you can make a very very very large, shallow black resevoir to evaporate water in.
if you always get a negative reinforcement for an action, operant conditioning will cause the drivers to slow down. tickets and cops are not regular enough to train people to stop.
You give drivers too much credit (see also today's poll hehe). My experience is that veteran drivers still haven't figured out that most traffic lights outside of downtown areas are triggered by car sensors. They always manage to stop several feet in front of the white line with a sign saying "STOP BEHIND LINE ON RED", and therefore in front of the car sensor. They wait in vain for a green light that will not come until a car pulls up behind them to trip the sensor long enough. On feeder roads to busy highways, the timers are usually either very long or disabled at off-peak times. The funny part is that near my neighborhood, it's always they same people waiting in the crosswalk... inching up every few seconds.
The lower temperature and reduced solar wind definately help the sounds carry further. It's also worth noting that train engines and their massive weight on tracks produce intense low frequency sounds that are nondirectional by nature.
Train engines are 16 to 18 cylinders, 10-15 liters per cylinder and usually run at around 1000 RPM.
By observations, church pipe organs seem to be able to produce some pretty low & loud notes. I don't recall seeing any chambers like this in the cathedrals I've visited
That is because pipe organs produce the sound by vibrating the air in the pipe - which you will notice is very long and usually several times wider than the pipes on the other side of the chromatic scale.
The future of Gateway looks pretty bleak. I think in this move they are probably looking to be acquired.
In some ways this is a reverse merger. The eMachines CEO and founder is replacing Gateway founder Ted Waitt. The headquarters are moving from suburban San Diego to Orange County, CA near eMachine's current Irvine, CA headquarters. It looks like the visonaries and operational efficiencies of eMachines are pushing to the top to save the company.
This highlights a fact that the previous poster missed. The architectures of the Pentium 4 and Itanium families are vastly different from the "x86 architecture" in the previous generations of chips. The x86 instruction set is there and the chip presents itself in a backwords compatible way to the system. The innards are vastly evolved, however. A similar analogy is the leap from the Pentium Pro/II/III architecture or K6 architecture to the Athlon architecture.
SPARC, MIPS, and PA-RISC have had relatively minor architecture changes over the same time period. The IBM Power chips have had much better evolutionary gains.
Or, in other words, we have a nice duopoly (Optus and Telstra, with Telstra doing most of the running these days, it would appear)
What about PowerTel's entry in to the market? I know they provide a feed to Macquarie (in adition to Optus) for example. Still, the biggest complaint I hear from IT peers downunder is the lack of choice for providers.
The problem with your theory is there was a massive increase in consumer bandwidth between 1996 and 2000.
No, I mentioned that few people had high speed Internet access in 1996. Do you think broadband will stop improving? Cable Internet in my area just doubled its speed. Quality, on the other hand, does introduce a problem in the theory - somehow most shows are not the same in a 320x200 format stretched to 1024x768 or larger... but innovations in compression combined with bandwidth increases will help. Besides, do you have to drop everything you're doing to download a few television shows? No. You can even watch an episode now while the next one downloads.
It's debatable whether a few people downloading episodes of their favorite TV programs can significantly impact the entertainment industry.
That is what people said about MP3s back in 1996 - the few people who were aware of their existance. Mostly college students had access to high speed connections, Napster didn't exist yet, and WinAMP was not released for another year (remember WinPlay3?).
Now people selling entertainment are under presure to prevent their business model slipping out from under them - this is why we see DRM being added to HDTV, for example. The question is, just like the CD purchasing vs. MP3 downloading story posted today on Slashdot, whether people downloading these shows are likely to fit in to one of three categories:
* Already subscribe to the channel/show pirated * Would not subscribe to the channel/show pirated unless it were free * Would subscribe to the channel/show pirated if it were not available free
The answer is going to be some of each, but probably more of the first two categories than the latter.
I just chose 5% because that is approximately the lowest rate businesses can borrow at. Inflation has been around 3.5% since 1913. If I chose the 10% average of equities then people would complain it is an unlikely rate. Thinking of inflation, I also could have applied inflation to the 10 year renewals, resulting in a smaller difference.
At $10/yr they aren't saving money... And if you consider the value of that $1,000 invested in other ways or used to pay down interest:
$14.99 per year invested at 5% compounded annually for 100 years is worth $41,080.49 in 100 years. $1,000 invested up front at 5% compounded annually for 100 years is worth $131,501.26 in 100 years.
The $14.99 is the per year rate for 10 year renewals on Network Solutions. Since Network Solutions sends dozens of emails reminding you to renew, you have no reason to forget. Just make sure the contact address is not an individual but hostmaster@ or domainadmin@ since a 10 year tenure at a company is rare these days. I think time is unlikely to account for this difference in opportunity cost.
You must be looking at the wrong powerbook. Apple's store won't let me look at the one you configured, but I definately just configured a $2723.00 with 1.25GB RAM, 80GB hard disk, AirPort Extreme and Super Drive and 12.1" display. Then I compared that to all of the $950 to $1000 laptops listed in the value section of Buy.com, most of which have 12" or 14" screens, similar form factors and weights (4.8lbs for this apple with upgrades inside). The battery life I can definately substantiate. Non-Centrino Pentium IV and Pentium III-M laptops in my real world trials, runnning the battery dry, have substantially shorter battery life than PowerPC laptops, but the Pentium IV Centrinos certainly have better performance than a 1GHz G4.
The Apple costs 3 times more than the Dell. Do you have 3 times more stuff?
If you recall the parent post, I was only refuting the claims that a $3,000 Apple computer is equivalent to a $1,000 PC laptop. I never said the Apple has three times the "stuff." Your Inspiron 5100 is a very different form factor from the PowerBook I used, but to compare to the PowerBook in performance features it will cost $1896 (1GB RAM, 60GB HDD, XP Pro, 812.11g card, DVD burner). The Inspiron 300m is closer in size, but without an internal DVD/CDROM drive. Upgrade it to XP Pro, 1.1GB RAM, 60GB, 802.11g, external DVD-R/CD-R and the price is $2,356. That's a really nice laptop, but it doesn't look like a 3x price difference to me.
Traditional air conditioning assumes cheap electricity, and plenty of peak capacity. (Ever try to start a compressor motor?)
I wonder if screw compressors defeat this problem as well? Advantages of screw compressors are the decreased noise and increased efficiency. Reciprocating compressors lose energy with inertia changes between pumpings (also where most of the noise comes from). The initial pumping is what takes so much energy to get started on reciprocating compressors. A screw compressor can start slow and allow coolant to bypass the compressor, then slowly allow more coolant to be compressed. I imagine someone could adapt this design to off grid power sources if they aren't already well suited for them.
Electric cars, maybe, although hybrids seem to be doing pretty well with plain old lead-acid batteries.
Lead-acid batteries are considered to heavy for most hybrids, since every bit of weight counts, especially when you have a weaker powerplant. The Toyota, Honda and Ford hybrid vehicles use sealed nickel-metal hydride (Ni-MH).
I compared it to the price of Microdrives and other compact flash form hard disks, not compact flash RAM. Sorry for the confusion. The drive inside is a 4GB Hitachi Microdrive.
And don't forget the laptops either. Honestly the only thing the $1000 laptop will have over a fully loaded, $2,700 PowerBook is a two button touchpad. The PowerBook will have 5 times the RAM, twice the battery life, over 2.5 times the storage capacity, a DVD/CD burner (a few sub $1,000 PC laptops are starting to include these now), 802.11g over 802.11b, and bluetooth vs. SIR IrDA if you're lucky.
Re:Yes, yes, yes, Apple's dying, blah blah blah
on
Why iPod Can't Save Apple
·
· Score: 2, Insightful
Ahh, but see, you're on to something there. For one, it is worth noting that the retail price of a compact flash card with the same capacity of the iPod Mini (which uses an OEM version of such a card) is greater than the retail price of the iPod Mini. The $20 engraving, which actually is about what you'd pay at your local mall, is where they make some margins back since it's less expensive if it is part of the assembly process. The $40 headphones? There are some decent margins there, but considering similar (quality and design) headphones from Sony are about $35, not as much as you might think. Still, the accessories and addons, and of course the iTunes store, are where the money comes from.
Slashdot Mirror
Person A receives an allegedly unsolicited email from you.
Person A complains to Spam Cop.
Spamcop tells you "someone complained. we're getting you shut down."
You say "What was his email address? I have the logs of the date, time, and IP address from which he asked to receive this newsletter. If through some never-before-seen miracle he got on my mailing list without signing up for it, I'll be more than happy to take him off."
Spamcop says, "Sorry, that would make it too easy to prove that it's this idiot's fault for asking to receive special offers."
You say "fine, I'll just remove him from the list. What's his email address?"
spamcop says, "sorry, we'd much rather stop all bulk emailing than actually solve individual's problems. FOAD."
You missed the very important part about the complaint being sent from an alias to the email address of the complainer. SpamCop does not mediate in your hypothetical transaction. The spammer and/or the ISP both have the opportunity to reply to this user. If the user does not give a valid response, in this case including the original address, the ISP and spammer have no reason to change a thing. SpamCop can't automagically shut people down, and ISPs generally only consider taking action after getting large numbers of SpamCop and other abuse reports.
If you run a newsletter, some flakey users quickly forget they signed up for it, start complaining via SpamCop, and then never reply to your response to their contant spam reports, you can contact SpamCop and request to refuse reports with the email address removed or refuse SpamCop reports alltogether. SpamCop will even block users who abusivly submit reports of questionable quality.
It's true that local delivery is the MDA's job, but a powerful MDA is built in to Postfix to take care of this via maildir, mbox, pipe or LMTP (for handing off to Cyrus, for example). Sendmail, however, requires an external MDA and seperate configuration to do what can be configured in a base install of Postfix. That's what's great about Postfix - it isn't just an MTA, but a mail system. Best of all, its features are well integrated but seperate processes to make them highly efficient.
I agree. I'm guessing that at least some of these complaints were emails forwarded to uce@ftc.gov, which receives millions of emails. The FTC has to sift through these, figure out which ones are worth going after, then figuring out which ones are from the same spammer, and finally which emails are usefull in a court case.
It is also important to note that Postfix provides Maildir support for local delivery. This means you can have nested folders (containing both messages and more folders) on your IMAP server, where as with Sendmail's mbox format you can only have folders containing messages, and those folders are actually just long text files. Qmail provides the maildir format natively, but Postfix makes it free.
this is, as far as i recall, the first time that apple has offred a video card option in its laptops.
It is not an option. If you click on one of the links so generiously provided in the post, such as the PowerBooks link, you can see that the option is either get a 15" or 17" PowerBook with an ATI Mobility Radeon 9700 or a 12" PowerBook with an NVIDIA GeForce FX Go5200.
Surely someone with a fast CPU on a fast connection is worth more per "CPU hour" than someone with a 486 on dialup?
The program probably is limited by IO, so the process sleeps once the outbound mail queue fills up. Translation: if you can only send 100 mails per hour on a Pentium 200 because of your bandwith limitation, the process will use just as much CPU as if the Pentium 200 were sending 100 emails in a few seconds over a 1.5Mbps link. However, you're right that an hour of CPU time on a 486 is worth much less than an hour of CPU time on an Athlon FX-53.
I don't know if anyone's tried scaling it. I imagine that for densely populated places that demand very large amounts of drinkable water on a daily basis, it would require impossible amounts of evaporation surface.
According to this page, the low end of the scale is about 11kWhr per 1,000 gallons (3,785L) for reverse osmosis. The Tampa plant produces up to 111,000,000 gallons per day. So that comes to 1,221,000 kWhr per day. We can skip the electrical conversion and use solar heat to evaporate and distill the water, but that requires much more power than reverse osmosis. I'm going to have to say this means it won't scale well, except perhaps in desert locations where you can make a very very very large, shallow black resevoir to evaporate water in.
if you always get a negative reinforcement for an action, operant conditioning will cause the drivers to slow down. tickets and cops are not regular enough to train people to stop.
You give drivers too much credit (see also today's poll hehe). My experience is that veteran drivers still haven't figured out that most traffic lights outside of downtown areas are triggered by car sensors. They always manage to stop several feet in front of the white line with a sign saying "STOP BEHIND LINE ON RED", and therefore in front of the car sensor. They wait in vain for a green light that will not come until a car pulls up behind them to trip the sensor long enough. On feeder roads to busy highways, the timers are usually either very long or disabled at off-peak times. The funny part is that near my neighborhood, it's always they same people waiting in the crosswalk... inching up every few seconds.
The lower temperature and reduced solar wind definately help the sounds carry further. It's also worth noting that train engines and their massive weight on tracks produce intense low frequency sounds that are nondirectional by nature.
Train engines are 16 to 18 cylinders, 10-15 liters per cylinder and usually run at around 1000 RPM.
Some locomotive engine specifications:
* GE Transportation Systems
* GM's Electro-Motive Division
By observations, church pipe organs seem to be able to produce some pretty low & loud notes. I don't recall seeing any chambers like this in the cathedrals I've visited
That is because pipe organs produce the sound by vibrating the air in the pipe - which you will notice is very long and usually several times wider than the pipes on the other side of the chromatic scale.
The future of Gateway looks pretty bleak. I think in this move they are probably looking to be acquired.
In some ways this is a reverse merger. The eMachines CEO and founder is replacing Gateway founder Ted Waitt. The headquarters are moving from suburban San Diego to Orange County, CA near eMachine's current Irvine, CA headquarters. It looks like the visonaries and operational efficiencies of eMachines are pushing to the top to save the company.
This highlights a fact that the previous poster missed. The architectures of the Pentium 4 and Itanium families are vastly different from the "x86 architecture" in the previous generations of chips. The x86 instruction set is there and the chip presents itself in a backwords compatible way to the system. The innards are vastly evolved, however. A similar analogy is the leap from the Pentium Pro/II/III architecture or K6 architecture to the Athlon architecture.
SPARC, MIPS, and PA-RISC have had relatively minor architecture changes over the same time period. The IBM Power chips have had much better evolutionary gains.
Or, in other words, we have a nice duopoly (Optus and Telstra, with Telstra doing most of the running these days, it would appear)
What about PowerTel's entry in to the market? I know they provide a feed to Macquarie (in adition to Optus) for example. Still, the biggest complaint I hear from IT peers downunder is the lack of choice for providers.
The problem with your theory is there was a massive increase in consumer bandwidth between 1996 and 2000.
No, I mentioned that few people had high speed Internet access in 1996. Do you think broadband will stop improving? Cable Internet in my area just doubled its speed. Quality, on the other hand, does introduce a problem in the theory - somehow most shows are not the same in a 320x200 format stretched to 1024x768 or larger... but innovations in compression combined with bandwidth increases will help. Besides, do you have to drop everything you're doing to download a few television shows? No. You can even watch an episode now while the next one downloads.
Capitalism at it's best.
It's called a captive market.
It's debatable whether a few people downloading episodes of their favorite TV programs can significantly impact the entertainment industry.
That is what people said about MP3s back in 1996 - the few people who were aware of their existance. Mostly college students had access to high speed connections, Napster didn't exist yet, and WinAMP was not released for another year (remember WinPlay3?).
Now people selling entertainment are under presure to prevent their business model slipping out from under them - this is why we see DRM being added to HDTV, for example. The question is, just like the CD purchasing vs. MP3 downloading story posted today on Slashdot, whether people downloading these shows are likely to fit in to one of three categories:
* Already subscribe to the channel/show pirated
* Would not subscribe to the channel/show pirated unless it were free
* Would subscribe to the channel/show pirated if it were not available free
The answer is going to be some of each, but probably more of the first two categories than the latter.
I just chose 5% because that is approximately the lowest rate businesses can borrow at. Inflation has been around 3.5% since 1913. If I chose the 10% average of equities then people would complain it is an unlikely rate. Thinking of inflation, I also could have applied inflation to the 10 year renewals, resulting in a smaller difference.
At $10/yr they aren't saving money... And if you consider the value of that $1,000 invested in other ways or used to pay down interest:
$14.99 per year invested at 5% compounded annually for 100 years is worth $41,080.49 in 100 years.
$1,000 invested up front at 5% compounded annually for 100 years is worth $131,501.26 in 100 years.
The $14.99 is the per year rate for 10 year renewals on Network Solutions. Since Network Solutions sends dozens of emails reminding you to renew, you have no reason to forget. Just make sure the contact address is not an individual but hostmaster@ or domainadmin@ since a 10 year tenure at a company is rare these days. I think time is unlikely to account for this difference in opportunity cost.
You must be looking at the wrong powerbook. Apple's store won't let me look at the one you configured, but I definately just configured a $2723.00 with 1.25GB RAM, 80GB hard disk, AirPort Extreme and Super Drive and 12.1" display. Then I compared that to all of the $950 to $1000 laptops listed in the value section of Buy.com, most of which have 12" or 14" screens, similar form factors and weights (4.8lbs for this apple with upgrades inside). The battery life I can definately substantiate. Non-Centrino Pentium IV and Pentium III-M laptops in my real world trials, runnning the battery dry, have substantially shorter battery life than PowerPC laptops, but the Pentium IV Centrinos certainly have better performance than a 1GHz G4.
The Apple costs 3 times more than the Dell. Do you have 3 times more stuff?
If you recall the parent post, I was only refuting the claims that a $3,000 Apple computer is equivalent to a $1,000 PC laptop. I never said the Apple has three times the "stuff." Your Inspiron 5100 is a very different form factor from the PowerBook I used, but to compare to the PowerBook in performance features it will cost $1896 (1GB RAM, 60GB HDD, XP Pro, 812.11g card, DVD burner). The Inspiron 300m is closer in size, but without an internal DVD/CDROM drive. Upgrade it to XP Pro, 1.1GB RAM, 60GB, 802.11g, external DVD-R/CD-R and the price is $2,356. That's a really nice laptop, but it doesn't look like a 3x price difference to me.
Traditional air conditioning assumes cheap electricity, and plenty of peak capacity. (Ever try to start a compressor motor?)
I wonder if screw compressors defeat this problem as well? Advantages of screw compressors are the decreased noise and increased efficiency. Reciprocating compressors lose energy with inertia changes between pumpings (also where most of the noise comes from). The initial pumping is what takes so much energy to get started on reciprocating compressors. A screw compressor can start slow and allow coolant to bypass the compressor, then slowly allow more coolant to be compressed. I imagine someone could adapt this design to off grid power sources if they aren't already well suited for them.
Electric cars, maybe, although hybrids seem to be doing pretty well with plain old lead-acid batteries.
Lead-acid batteries are considered to heavy for most hybrids, since every bit of weight counts, especially when you have a weaker powerplant. The Toyota, Honda and Ford hybrid vehicles use sealed nickel-metal hydride (Ni-MH).
Many flywheel systems use smaller flywheels spinning much faster. Take this one, for example, at 60,000RPMS and 23Kg.
Still, if you can't afford lead acid batteries, you probably can't afford flywheels for the scale of the system.
The Ipod mini uses a hard disk.
I compared it to the price of Microdrives and other compact flash form hard disks, not compact flash RAM. Sorry for the confusion. The drive inside is a 4GB Hitachi Microdrive.
And don't forget the laptops either. Honestly the only thing the $1000 laptop will have over a fully loaded, $2,700 PowerBook is a two button touchpad. The PowerBook will have 5 times the RAM, twice the battery life, over 2.5 times the storage capacity, a DVD/CD burner (a few sub $1,000 PC laptops are starting to include these now), 802.11g over 802.11b, and bluetooth vs. SIR IrDA if you're lucky.
Ahh, but see, you're on to something there. For one, it is worth noting that the retail price of a compact flash card with the same capacity of the iPod Mini (which uses an OEM version of such a card) is greater than the retail price of the iPod Mini. The $20 engraving, which actually is about what you'd pay at your local mall, is where they make some margins back since it's less expensive if it is part of the assembly process. The $40 headphones? There are some decent margins there, but considering similar (quality and design) headphones from Sony are about $35, not as much as you might think. Still, the accessories and addons, and of course the iTunes store, are where the money comes from.